For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Python class

SymbolicDim

SymbolicDim

class max.graph.SymbolicDim(value)

source

Bases: Dim

A symbolic tensor dimension with an unknown size identified by name.

When you don’t know a dimension value at compile time, you can use a symbolic dimension. This helps you identify dimensions by name and lets the compiler optimize operations when two or more dimensions share the same name.

The following example creates a symbolic dimension implicitly passing the strings "batch" and "x" to TensorType:

from max.dtype import DType
from max.graph import DeviceRef, TensorType

tensor_type = TensorType(DType.float32, ("batch", "x", 10), device=DeviceRef.CPU())

Converts valid input values to Dim.

Parameters:

name (str)

from_mlir()

static from_mlir(attr)

source

Constructs a SymbolicDim from a kgen.ParamDeclRefAttr.

Parameters:

attr (TypedAttr) – The kgen.ParamDeclRefAttr to parse into a SymbolicDim.

Returns:

The SymbolicDim represented by the kgen.ParamDeclRefAttr.

Return type:

SymbolicDim

name

name: str

source

The name of the dimension.

parameters

property parameters: Iterable[SymbolicDim]

source

Lists the symbolic dimension names on which this dim depends.

substitute()

substitute(mapping)

source

Returns this dim with named symbols replaced per mapping.

Parameters:

mapping (Mapping[str, int | str | Dim | integer | TypedAttr]) – A mapping from symbolic dimension name to the replacement dim (or dim-like value).

Returns:

Dim(mapping[self.name]) if self.name is in mapping, otherwise this dim unchanged.

Return type:

Dim

to_mlir()

to_mlir()

source

Creates an mlir.Attribute representing this dimension.

This is used internally when constructing tensor MLIR types.

Returns:

An mlir.Attribute in the context representing the dimension.

Return type:

ParamDeclRefAttr